arXiv subjects

Yiheng Zhou

Publications and source records attributed to Yiheng Zhou.

6 recordsLinked to original sources

arXiv2026

Unlocking Model Potentials Through Adaptive Multi-Agent Scaffolding for Efficient Issue Resolution

Resolving issues with ambiguous and incomplete descriptions, particularly concerning complex bugs, requires a sophisticated, long-horizon workflow. Agents must navigate codebases to locate the root cause, reproduce the failure, implement a fix, and validate the resulting patch. Inefficient context management, thereby, can lead to rapid context degradation and context poisoning, preventing successful resolution. We propose icat-agent, a decentralized, multi-agent scaffolding that replaces shared context with synchronous, event-based message passing. Utilizing a rubric-based issue quality check, icat-agent strategically pivots its workflow: it initiates parallel patching and validation for well-defined issues, while deploying preliminary exploration for low-quality ones. A comprehensive evaluation of icat-agent on SWE-bench Verified and SWE-bench Pro demonstrates that it consistently outperforms prominent baselines across all difficulty levels, including SWE-agent, mini-SWE-agent, and Claude Code, while using the same underlying models, improving by 3.6-8.4% on SWE-bench Verified and 6.3-18.5% on SWE-bench Pro. icat-agent is also computationally efficient, reducing the average cost by $1.18 per instance compared with the multi-agent Claude Code baseline. Our findings reveal that a robust scaffold such as icat-agent unlocks substantial latent capability within a fixed model, with the same backbone resolving markedly more issues under icat-agent than under existing scaffolds. icat-agent +GPT-5.4-xhigh resolves 67.4% of SWE-bench Pro problems, outperforming the current best result on SWE-bench Pro (59.10%, mini-SWE-agent+GPT-5.4-xhigh) by 8.3 percentage points.

cs.SE↗

arXiv2023

On Orderings of Probability Vectors and Unsupervised Performance Estimation

Unsupervised performance estimation, or evaluating how well models perform on unlabeled data is a difficult task. Recently, a method was proposed by Garg et al. [2022] which performs much better than previous methods. Their method relies on having a score function, satisfying certain properties, to map probability vectors outputted by the classifier to the reals, but it is an open problem which score function is best. We explore this problem by first showing that their method fundamentally relies on the ordering induced by this score function. Thus, under monotone transformations of score functions, their method yields the same estimate. Next, we show that in the binary classification setting, nearly all common score functions - the $L^\infty$ norm; the $L^2$ norm; negative entropy; and the $L^2$, $L^1$, and Jensen-Shannon distances to the uniform vector - all induce the same ordering over probability vectors. However, this does not hold for higher dimensional settings. We conduct numerous experiments on well-known NLP data sets and rigorously explore the performance of different score functions. We conclude that the $L^\infty$ norm is the most appropriate.

cs.LG↗

arXiv2019

Augmenting Non-Collaborative Dialog Systems with Explicit Semantic and Strategic Dialog History

We study non-collaborative dialogs, where two agents have a conflict of interest but must strategically communicate to reach an agreement (e.g., negotiation). This setting poses new challenges for modeling dialog history because the dialog's outcome relies not only on the semantic intent, but also on tactics that convey the intent. We propose to model both semantic and tactic history using finite state transducers (FSTs). Unlike RNN, FSTs can explicitly represent dialog history through all the states traversed, facilitating interpretability of dialog structure. We train FSTs on a set of strategies and tactics used in negotiation dialogs. The trained FSTs show plausible tactic structure and can be generalized to other non-collaborative domains (e.g., persuasion). We evaluate the FSTs by incorporating them in an automated negotiating system that attempts to sell products and a persuasion system that persuades people to donate to a charity. Experiments show that explicitly modeling both semantic and tactic history is an effective way to improve both dialog policy planning and generation performance.

cs.CL↗

arXiv2019

A Dynamic Strategy Coach for Effective Negotiation

Negotiation is a complex activity involving strategic reasoning, persuasion, and psychology. An average person is often far from an expert in negotiation. Our goal is to assist humans to become better negotiators through a machine-in-the-loop approach that combines machine's advantage at data-driven decision-making and human's language generation ability. We consider a bargaining scenario where a seller and a buyer negotiate the price of an item for sale through a text-based dialog. Our negotiation coach monitors messages between them and recommends tactics in real time to the seller to get a better deal (e.g., "reject the proposal and propose a price", "talk about your personal experience with the product"). The best strategy and tactics largely depend on the context (e.g., the current price, the buyer's attitude). Therefore, we first identify a set of negotiation tactics, then learn to predict the best strategy and tactics in a given dialog context from a set of human-human bargaining dialogs. Evaluation on human-human dialogs shows that our coach increases the profits of the seller by almost 60%.

cs.CL↗

arXiv2016

Fine-grained Mining of Illicit Drug Use Patterns Using Social Multimedia Data from Instagram

According to NSDUH (National Survey on Drug Use and Health), 20 million Americans consumed drugs in the past few 30 days. Combating illicit drug use is of great interest to public health and law enforcement agencies. Despite of the importance, most of the existing studies on drug uses rely on surveys. Surveys on sensitive topics such as drug use may not be answered truthfully by the people taking them. Selecting a representative sample to survey is another major challenge. In this paper, we explore the possibility of using big multimedia data, including both images and text, from social media in order to discover drug use patterns at fine granularity with respect to demographics. Instagram posts are searched and collected by drug related terms by analyzing the hashtags supplied with each post. A large and dynamic dictionary of frequent drug related slangs is used to find these posts. User demographics are extracted using robust face image analysis algorithms. These posts are then mined to find common trends with regard to the time and location they are posted, and further in terms of age and gender of the drug users. Furthermore, by studying the accounts followed by the users of drug related posts, we extract common interests shared by drug users.

cs.SI↗

arXiv2016

Understanding Illicit Drug Use Behaviors by Mining Social Media

Drug use by people is on the rise and is of great interest to public health agencies and law enforcement agencies. As found by the National Survey on Drug Use and Health, 20 million Americans aged 12 years or older consumed illicit drugs in the past few 30 days. Given their ubiquity in everyday life, drug abuse related studies have received much and constant attention. However, most of the existing studies rely on surveys. Surveys present a fair number of problems because of their nature. Surveys on sensitive topics such as illicit drug use may not be answered truthfully by the people taking them. Selecting a representative sample to survey is another major challenge. In this paper, we explore the possibility of using big data from social media in order to understand illicit drug use behaviors. Instagram posts are collected using drug related terms by analyzing the hashtags supplied with each post. A large and dynamic dictionary of frequent illicit drug related slang is used to find these posts. These posts are studied to find common drug consumption behaviors with regard to time of day and week. Furthermore, by studying the accounts followed by the users of drug related posts, we hope to discover common interests shared by drug users.

cs.SI↗